Biological characteristic authentication is a sort of identity authentication using human biological characteristics, such as voice, fingerprint, iris and signature. As an accurate, fast and efficient identity authentication, biological characteristic authentication is becoming more and more widely used in various fields that require authentication. Amongst biological characteristic authentications, face identity authentication which is based on face recognition is one of the most popular biological characteristic authentication.
Conventional face identity authentication methods collect standard face picture, extract the feature data from the standard face picture and store the feature data as the template in a database in an authentication server. When it is required for identity authentication, feature data is extracted from an inputted face picture, and subsequently compared with the template that is previously stored in the database. If the feature data matches the template, the authentication is passed, otherwise the authentication fails.
When the conventional face identity authentication is applied to the internet, the internet client and the authentication server are not in an equally real environment, the authentication server cannot distinguish whether the inputted face picture or video is acquired exactly from a real person or a duplicated fake that is taken previously. A malicious third party may input a face picture or video taken previously to deceive the authentication server, and if the picture or video pass the authentication, users' accounts and virtual property will be robbed, resulting in great risk to the security of users' internet applications.
Therefore, heretofore unaddressed needs exist in the art to address the aforementioned deficiencies and inadequacies.